Here is a summary of this blog post:

• Automated Decision Making has been successful for our customers.
• Our customers primarily use AI as a labor assistance tool.
• AI (specifically LLMs) as a feature within a product has not yielded success for our customers.
• Read on for more details

Platforms like the Internet or the smartphone were the birthplace of startup booms. AI, however, is not as straightforward. While its capabilities to generate code and media could lead to another startup boom, we are not yet seeing that level of automation. Here are ways our customers leverage “Automated Decision Making” and AI in their businesses.

Automated Decision Making

Traditional ADM focuses on making high-scale business decisions. A business process begins with a business analysis—such as inventory analysis, identifying best sellers, product profitability, etc.—which leads to a decision, like prioritizing the sale of certain products or discontinuing others. If this process is regularly repeated, it can likely be automated through ADM. ADM's role is to automate both the decision-making process and the actions resulting from the analysis. Let's explore a concrete example.

Consider a business team tasked with creating a “recommended products” list for display below a product page. The initial approach might involve analyzing all past orders to create a “Frequently Bought Together” list. Once the analysis is complete, these recommended products are uploaded as a section on each product page. How often should these recommendations be updated? While Amazon is known for updating recommendations within minutes of behavioral changes, a small shop might only update this list annually.  

For another example, please refer to our case study on Artificial Intelligence. ADM can be applied in any industry and hinges on a business decision followed by an analysis.

Automated Decision Making scales the work of business analysts and the subsequent actions from their analyses, hence the term “decision making.” It has the capability to both analyze and make decisions.

Artificial intelligence

Artificial intelligence is primarily used as a smart assistant. It doesn’t make decisions unless integrated with a custom integration. A custom integration with an LLM, like a chatbot feature, can perform tasks within your app. However, this hasn't been particularly useful, as consumers often prefer logging into their dashboards to make changes directly. Consequently, few customers see a business incentive to include it in their products.

Companies that have integrated ChatGPT as a “customer service” chatbot haven't seen significant benefits. It often frustrates customers seeking human interaction. Although a language interface can enhance product accessibility, we haven't identified a market that significantly benefits from it. Therefore, we currently do not recommend implementing ChatGPT in your product. That is all based on our experience, and can be corrected soon if we do happen to find a great use-case for it.

We personally utilize Artificial Intelligence to review code, provide guidance on system implementation, and synthesize essential information from documentation for patches or version upgrades. This makes us more efficient programmers compared to our competition. However, this advantage is not a lasting competitive edge, as this expertise will soon be expected in the workplace. We absolutely recommend using it within your business processes.

Conclusion

Since we don’t yet have fully automated AI decision-makers, ADM remains the most valuable tool for automating business intelligence and scaling your company. This may change in the future, but for now, consider integrating an LLM (such as ChatGPT, Claude, or Bard) within your teams to assist in answering questions. These tools can become essential for advising, directing, and training employees.